Rescued analog oscilloscope works fine, still has value

-February 28, 2017

A few weeks ago, my father-in-law called to say he found an oscilloscope. His neighbor was apparently going to put it in the trash so he asked if I wanted it. Of course I did. Who wouldn't want to save an oscilloscope from destruction? It's an LG OS-5020, 20 MHz, two-channel analog model in very good condition. Let's see how well it works.

The oscilloscope didn't have probes. Fortunately, I have two 150 MHz passive probes plus some BNC cables to connect it to a function generator. The oscilloscope powered up (no boot time like a digital oscilloscope). I set it to CH1, adjusted the knobs and sure enough, a trace appeared. Now it was time to connect it to some signals and get out a lab notebook. Figure 1 shows the test setup, which consists of an HP34401A DMM and a National Instruments VirtualBench (computer for VirtualBench not shown).

Figure 1 The oscilloscope under test, an HP34401A DMM, and a National Instruments VirtualBench (PC not shown) served as the test setup.

For a first test, I went for the basics, DC and a 9V battery. Measuring the voltage on the DMM showed the battery at 8.368V. Connecting a probe to the battery showed a DC level (Figure 2) of what I'd call "pretty close" to that level.

Figure 2 The LG oscilloscope shows a DC level when connected to a battery. The level is reasonably close to the value measured with a DMM.

So far, so good. Next, I connected a probe to the LG's calibration point, a 1 kHz square wave with an amplitude of 0.5 VP-P. Here's a photo of the screen, taken with my phone (Figure 3). Yes, that's the only way to get a screen image on an analog oscilloscope. At least I didn't have to use a bulky Polaroid camera to get black-and-white photos like I once did.

Figure 3 The oscilloscope screen shows its 1 kHz calibration signal.

The top parts of the signal are somewhat skewed and that skew differs from left to right. Indeed, the falling edges appear to occur in negative time. For a sanity check, I connected the VitrualBench oscilloscope to the calibration point (Figure 4).

Figure 4 The LG oscilloscope's calibration signal looks properly aligned when viewed with the digital VirtualBench.

Correcting the "negative time" was easy. The oscilloscope's trace needed to be rotated. If you're not familiar with analog oscilloscopes, remember that the trace on the screen comes from a beam that needs to align with the grid. That's why this instrument has a trace rotation adjustment. Digital oscilloscopes don't need that because there's no beam, just illuminated pixels.

Next, I set the VirtualBench function generator to produce a sine wave at 1 kHz, then compared the LG screen to the VirtualBench at frequencies up to 20 MHz, the function generator's highest frequency. Figure 5 shows the LG screen with a 10 VP-P, 1 kHz sine wave (0.5 µs/div, 2 V/div, 1× probe). For the remainder of the tests, I connected the VirtualBench function generator to the LG oscilloscope through a BNC cable.

Figure 5 A 10 VP-P, 1 kHz sine wave displays cleanly on the LG analog oscilloscope.

That sine wave compares well to the VirtualBench display (Figure 6).

Figure 6 The 1 kHz sine wave shown on the VirtualBench display shows that the analog oscilloscope is in pretty good shape.

Next, I set the VirtualBench function generator's frequency to 20 MHz, still at 10 VP-P. In Figure 7, you can see the drop in amplitude at the analog oscilloscope's maximum frequency. Not shown are photos of the screen at frequencies of 1 MHz, 2 MHz, 5 MHz, 10 MHz, 15 MHz, 17 MHz, 18 MHz, and 19 MHz.

Figure 7 At it's full rated bandwidth, the LG analog oscilloscope shows the expected decreased sine wave's amplitude.

At 20 MHz, you can see the difference that oscilloscope bandwidth makes. In Figure 8, you can see the higher-bandwidth VirtualBench's display of the 20 MHz sine wave.

Figure 8 With a 20 MHz signal, the 100 MHz VirtualBench display doesn’t show the decreased amplitude seen on the 20 MHz LG oscilloscope.

Note: The VirtualBench oscilloscope's bandwidth is 100 MHz and I had a 150 MHz passive probe. No problem? Not quite. The probe's bandwidth at the 1× setting is rated at 6 MHz. Be aware of a probe's greatly reduced bandwidth at the 1× setting. I used a BNC cable for most of these tests.

Just for fun, I tried a 5V, 1 MHz square wave (Figure 9). Both oscilloscopes showed some ringing on the signal's rising and falling edges, possibly caused by parasitic capacitance on the connecting cable or perhaps an impedance mismatch from the function generator's output to the oscilloscope's input.

Figure 9 A 5 MHz square wave has ringing on the rising and falling edges – on both the analog LG and digital VirtualBench oscilloscopes.

Now what?

Given that the LG oscilloscope was headed for destruction, it was a good save. It could use a calibration, but that's about all.

While keeping or selling the LG OS-5020 analog oscilloscope are options, I prefer to give it to a school or nonprofit maker space. I have contacted a local maker space and hopefully, they will take it. Check the comments for updates.

Lastly, here are two pages from my lab notebook (Figure 10), complete with glossy printed photos of oscilloscope screens: analog signal to digital photo to analog photo to digital photo.

Figure 10 Printed photos of oscilloscope screens was once the way to document measurements. Today, we just take screen images and paste them into word processors.

Martin Rowe covers test and measurement for EDN and EE Times. Contact him at Circle me on Google+ Follow me on TwitterVisit my LinkedIn page

Also see:


Loading comments...

Write a Comment

To comment please Log In